The Office for Budget Responsibility’s (OBR) March 2026 Economic and Fiscal Outlook contains seven direct references to artificial intelligence — not many for a 200-page document. But those seven references carry a combined fiscal weight of £90 billion. That is the gap between the OBR’s optimistic AI-driven productivity scenario (borrowing £50 billion lower by 2030-31) and its downside scenario where productivity stays flat (borrowing £40 billion higher). For UK businesses, this is not an abstract number. It is the difference between operating in a growing economy with rising tax receipts and a stagnant one where the Treasury looks for revenue wherever it can find it.

What the OBR actually assumes about AI

The OBR’s central forecast assumes productivity growth will rise to 1.0% by the end of the forecast period. That is double the 0.5% annual growth the UK has averaged since the financial crisis, but well below the 2% rates that were normal before 2008. Total factor productivity (TFP) — the measure of how efficiently the economy combines labour and capital — is forecast to increase gradually to 0.8% by 2030.

Within that TFP forecast, AI adoption is one of three factors the OBR expects to drive improvement, alongside the fading impact of past economic shocks and planning reforms introduced in March 2025. But the OBR does not quantify how much of the TFP improvement it attributes to AI specifically. It states that “the effects from AI specifically on productivity” are “highly uncertain” and points readers to a separate briefing paper for alternative scenarios.

Strategic Reality: The OBR treats AI as an upside risk factor, not a central planning assumption. UK government fiscal planning does not depend on AI delivering productivity gains — but the difference if it does is enormous.

This matters for business planning. If the OBR — the UK’s official fiscal watchdog — does not feel confident enough to bake AI productivity gains into its central forecast, businesses should be cautious about doing so in their own financial projections. The gains might materialise. The OBR thinks they could. But the range of uncertainty is wider than most boardroom discussions acknowledge.

The £90 billion swing: two very different Britains

The clearest AI-related insight in the report comes from the OBR’s productivity scenarios, originally developed in November 2025 and carried forward here:

Critical Context: A £90 billion swing in annual government borrowing by 2030-31 depending on productivity outcomes. For context, the entire UK defence budget is around £55 billion. AI-driven productivity is, by the OBR’s own modelling, one of the largest single variables in the UK’s fiscal future.

The upside scenario — 1.5% annual productivity growth — is what you get “due to a more optimistic scenario for the impact of AI.” That is the OBR’s own language. At that rate, borrowing falls by £50 billion a year and the UK’s debt trajectory looks substantially more manageable. The downside — 0.5% growth, matching the past 15 years — means £40 billion more in annual borrowing by the end of the decade.

Neither scenario is a prediction. Both are plausible. And that uncertainty is the point.

The “technological displacement” scenario nobody is discussing

Buried in Box 2.2 of the report, the OBR models a scenario it calls “technological displacement.” This is, in plain terms, the AI-takes-jobs scenario — and its conclusions should concern every HR director and workforce planner in the country.

In this scenario:

• New technology displaces workers and acts as a substitute for labour
• The equilibrium unemployment rate rises to 5.5% (from the central forecast of just over 4%)
• Productivity rises for workers who remain employed, fully offsetting the GDP impact of fewer people working
• But higher productivity is not reflected in higher real earnings — profits go up, wages do not
• The labour share of income falls while the corporate profit share rises

Hidden Cost: In the technological displacement scenario, GDP stays the same but the economy becomes fundamentally less equal. Higher profits, lower employment, flat wages. Government borrowing rises by £9 billion a year because personal tax and VAT revenues fall while welfare spending increases.

The fiscal arithmetic is telling. Even though GDP is unchanged in this scenario, borrowing rises by an average of £9 billion annually. Lower personal tax revenues (fewer people working) and higher welfare spending (more people claiming benefits) outweigh the higher corporation tax from increased profits. The government’s tax base becomes less “rich” because labour income faces a higher effective tax rate than corporate profits.

Compare this with the OBR’s “higher labour costs” scenario, where unemployment also rises to 5.5% but without any productivity offset. In that world, borrowing is £15 billion a year higher and underlying debt is 3.5% of GDP higher by 2030-31.

The message: AI-driven productivity gains are genuinely better for the public finances than the alternative — but they are not painless, and they do not pay for themselves in tax revenue terms if the gains flow predominantly to capital rather than labour.

What the global context tells us

The OBR notes that global GDP grew 3.3% in 2025, and that “headwinds from higher tariffs and uncertainty over trade policy are likely to have been broadly offset by higher investment, particularly in new technology such as artificial intelligence.” That is worth pausing on. Globally, AI investment is now acting as a counterweight to trade disruption.

For UK businesses, the implication is that AI investment is now a macroeconomic factor, not just a technology decision. When the IMF and OBR start treating AI capital expenditure as a measurable offset to tariff-related economic damage, it tells you something about the scale of capital flowing into AI infrastructure worldwide.

Competitive Reality: The UK’s central GDP growth forecast of 1.1% in 2026 lags behind global growth of around 3%. If other economies are investing in AI infrastructure at scale while UK productivity growth remains below 1%, the competitiveness gap widens with each passing year.

The UK’s own growth forecast — 1.1% in 2026, averaging 1.6% from 2027 to 2030 — sits in the lower half of the global distribution. This is partly structural (an ageing population, lower migration) and partly a consequence of weak productivity. The OBR projects that if pre-financial-crisis productivity trends had continued, UK GDP per person would be approximately 30% higher today. That is the cost of 15 years of stagnation, and AI is one of the few plausible routes to reversing it.

Five things UK businesses should take from this report

1. Do not build AI productivity gains into base-case financial forecasts

The OBR does not, and it has access to more economic modelling capability than any single business. Use the central forecast (1.0% productivity growth) as your baseline and treat AI-driven improvements as upside scenarios to be validated.

2. Plan for the workforce transition, not just the productivity gain

The technological displacement scenario is not the OBR’s most likely outcome, but it is one they consider plausible enough to model in detail. If your AI strategy assumes you will reduce headcount, you need a detailed plan for what happens to displaced workers — not just because it is the right thing to do, but because the fiscal consequences (higher welfare costs, lower tax receipts) will create political pressure for regulatory intervention.

Implementation Note: The OBR’s technological displacement scenario assumes the equilibrium unemployment rate rises to 5.5%. Businesses planning AI-driven restructuring should stress-test their plans against this labour market assumption, not the more benign central forecast of 4%.

3. Watch for tax policy shifts that follow the money

If AI does shift the income distribution from labour to capital — as the technological displacement scenario suggests — expect fiscal policy to follow. Higher corporation tax, windfall taxes on AI-driven profits, or changes to capital allowances are all plausible policy responses to a shrinking personal tax base. Build this regulatory risk into investment cases.

4. The investment window is now

Global AI investment is already offsetting trade disruption at a macroeconomic level. UK businesses that delay AI adoption are not standing still — they are falling behind in a context where competitors are using AI investment to absorb economic shocks. The OBR’s forecast suggests the UK economy is already underperforming, and the gap between AI-adopting and non-adopting firms will mirror the gap between the OBR’s upside and downside productivity scenarios.

5. Productivity measurement will matter more than ever

The OBR’s entire fiscal framework rests on productivity assumptions, and it admits these are “one of the most important, but uncertain, forecast judgements.” Businesses that can demonstrate measurable productivity gains from AI will be better positioned for potential tax incentives, capital allowances, or regulatory frameworks designed to reward adoption. Those that cannot measure their AI impact will not be able to make the case.

Take Action: Review your organisation’s productivity measurement framework now. Can you isolate AI’s contribution to output per hour? If not, you will struggle to justify continued investment when the next economic downturn compresses budgets.

The challenges the OBR does not address

For all its analytical depth, the report leaves several AI-related questions unanswered:

Sectoral distribution of AI impact. The OBR models productivity at an aggregate level. It does not break down which sectors will see AI-driven gains first, or which are most exposed to displacement. Financial services, professional services, and public administration are likely early movers, but the OBR offers no sectoral guidance.

Skills and retraining costs. The technological displacement scenario assumes workers lose jobs but does not model the fiscal cost of retraining them. If AI displaces workers in routine cognitive tasks, the transition costs could be substantial — and they would hit government spending at the same time as tax revenues fall.

AI safety and regulatory drag. The upside productivity scenario assumes AI adoption proceeds without major setbacks. A significant AI safety incident, a regulatory crackdown, or public backlash against automation could slow adoption and push the economy closer to the downside scenario.

International competitive dynamics. The OBR notes global AI investment but does not model how the UK’s AI adoption rate compares with key competitors. If US and Chinese firms gain productivity advantages faster, the UK’s relative position could deteriorate even if absolute productivity improves.

Reality Check: The OBR’s scenarios are useful bookends, but they are not forecasts. The actual path will depend on policy choices, business investment decisions, and technological developments that are not yet visible. The value of the analysis is in the range, not in any single number.

What this means for UK AI strategy

The OBR’s March 2026 forecast is, whether intended or not, the strongest quantified case any UK government body has made for the fiscal importance of AI adoption. A £90 billion range in annual borrowing — hinging substantially on whether productivity growth returns to pre-crisis levels, “for example, due to a more optimistic scenario for the impact of AI” — puts AI squarely at the centre of the UK’s medium-term economic outlook.

For business leaders, the takeaway is not that AI will definitely transform the economy. It is that the official fiscal watchdog considers AI-driven productivity to be one of the single largest variables in the UK’s economic future, whilst simultaneously admitting it has no confident way to predict the outcome.

That uncertainty is not a reason to wait. It is a reason to build organisations that can adapt to either scenario — investing in AI capability whilst maintaining workforce resilience, measuring productivity improvements rigorously, and preparing for a policy environment that will inevitably respond to whichever scenario begins to materialise.

The £90 billion question is not whether AI matters. It is whether UK businesses will be ready when the answer arrives.

Source: Office for Budget Responsibility, Economic and Fiscal Outlook — March 2026. Published March 2026.

Analysis by Resultsense — making sense of AI in the UK. For strategic guidance on AI adoption and implementation, explore our AI Strategy Blueprint and AI Implementation Support services.